Risk Narrative of Emergency and Disaster Management, Preparedness, and Planning (EDMPP): The Importance of the ‘Social’

Abstract

Risk perception, literacy, communication, narrative, governance, and education are important aspects of emergency and disaster management, preparedness, and planning (EDMPP) as they for example influence and direct EDMPP policies and actions. A thorough understanding of the ‘social aspects of risk is important for EDMPP, especially in relation to marginalized populations who are often overlooked. Technologies are increasingly employed for EDMPP. How these technology applications identify and engage with the ‘social’ of risk in general and the ‘social’ of risk experienced by marginalized populations is important for EDMPP. Equity, diversity, and inclusion (EDI) and similar phrases are employed as policy concepts to improve research, education, and participation in the workplace for marginalized groups such as women, Indigenous peoples, visible/racialized minorities, disabled people, and LGBTQ2S including in workplaces engaging with EDMPP which includes universities. The aim of this scoping review was to generate data that allows for a detailed understanding of the risk related discussions within the EDMPP academic literature as these discussions shape EDMPP policies and actions. The objective of this scoping review study was to map out the engagement with risk, specifically the social aspects of risk, in the EDMPP-focused academic literature with a focus on (a) EDMPP in general, (b) COVID-19, (c) EDMPP and marginalized groups, (d) EDMPP and patients, and (e) EDMPP and technologies (artificial intelligence, machine learning, machine reasoning, algorithm design approaches such as Bayesian belief networks, e-coaching, decision support systems, virtual coaching, automated decision support, e-mentoring, automated dialogue and conversational agents). Using the academic databases SCOPUS, Web of Sciences, and databases accessible under Compendex and EBSCO-HOST and performing hit count frequency searches of online and downloaded abstracts and thematic analysis of downloaded abstracts the study reveals a lack of coverage on the social aspects of risk and engagement with risk concepts such as risk perception, risk governance, risk literacy, risk communication, risk education and risk narrative especially in conjunction with marginalized groups and technologies employed in EDMPP decision support. Our findings suggest many opportunities to further the EDMPP academic inquiry by filling the gaps.

1. Introduction

Emergency and disaster management, preparedness, and planning (EDMPP) including COVID-19 is becoming ever more important [1,2,3,4,5,6]. Discussions around risk influence and direct EDMPP policies and actions [7,8,9,10,11,12,13,14]. Risk narratives are present in emergency and disaster policy documents [4,15] and marginalized groups are known to be at a higher risk of experiencing negative consequences of emergencies and disasters [4,15]. With that, it is important to understand the risk narrative within the EDMPP academic literature including how risk focused EDMPP academic literature understands and engages with the ‘social’ of risks, specifically in relation to marginalized groups. A recent scoping review looked at the general engagement with the ‘social’ in the EDMPP academic literature using a hit count approach [16]. The aim of this scoping review was to generate data on the EDMPP risk narrative and the ‘social’ using a hit count approach of online and downloaded abstracts and thematic analysis of downloaded abstracts allowing for a detailed understanding of the risk discussions within the EDMPP academic literature, which is a topic not covered in current academic literature [16]. Given the study’s aim, the first objective of this study was to generate frequencies for the terms “risk” and “social risk” mentioned in the EDMPP and e-coaching, decision support system, virtual coaching, automated decision support, e-mentoring, automated dialogue, and conversational agent focused academic literature in general and in relation to marginalized groups.

Risk discourses use many risk concepts in discussing emergencies and disasters such as risk perception [17,18,19,20], risk governance [21,22,23], risk literacy [24,25,26], risk communication [27,28,29,30,31,32,33,34,35], risk education [36,37,38] and risk narrative [39]. The “purpose of risk governance is to provide an understanding of how to with uncertain, complex, and/or unclear risks within a conceptual and normative basis” [40] (p. 431). Risk literacy has to improve in order to improve risk governance [41] and with that, risk education is seen as needed to increase risk literacy [42]. Risk narratives, and whether one uses the term risk are influenced by many factors [43,44]. Perception of risk is a social and cultural construct and is impacted by many factors [45]. Further, understanding “how risk perception becomes various depending on each position, value, and society” [41] (p. 1260) is important for developing risk literacy. How these concepts are employed in the EDMPP literature including technologies used in EDMPP impacts EDMPP actions and policies. Therefore, the second objective was to investigate using hit count searches of online and downloaded abstracts and a thematic analysis of downloaded abstracts the presence of the risk related concepts “risk governance”, “risk communication”, ”risk education”, “risk narrative” and “risk perception” in the literature covered and to investigate using a thematic analysis of downloaded abstracts what is said especially in relation to marginalized groups in conjunction with the risk concepts of “risk governance”, “risk communication”, ”risk education”, “risk narrative” and “risk perception”.

Many technologies are envisioned to be employed in EDMPP. At the same time, it is recognized that technologies contain risks including social risks and risk is a topic discussed within science and technology governance and ethics fields. Therefore, the third objective was to investigate the frequency of mentions in the literature we covered of science and technology governance terms and ethics fields in conjunction with the term risk.

Equity, diversity, and inclusion (EDI), other EDI phrases and EDI frameworks [46] are employed as policy concepts to improve research, education, and participation in the workplace for marginalized groups such as women, Indigenous peoples, visible/racialized minorities, disabled people, and LGBTQ2S which includes EDMPP linked research, education and workplaces. At the same time it is reported that marginalized people experience higher risks [4,15] and greater impact in relation to emergencies and disasters [47,48,49,50]. Therefore, the fourth objective was to investigate using hit count searches of online and downloaded abstracts and a thematic analysis of downloaded abstracts whether EDI concepts and frameworks as well as marginalized groups covered in EDI discourses are present in the risk narrative of the literature covered and how they are covered.

The study asked three questions: (1) How often are the terms “risk” and “social risk” mentioned in the literature covered (2) How often and how are the terms risk perception, risk governance, risk literacy, risk communication, risk education and risk narrative mentioned in the literature covered (3) How often and how is the term risk and the risk related concepts risk perception, risk governance, risk literacy, risk communication, risk education and risk narrative engaged with in the EDMPP literature in conjunction with EDI groups.

1.1. Risk Narratives and the COVID-19 Research Road Map

The COVID-19 research roadmap mentions risk many times such as “these knowledge needs are diverse, but they all illustrate a reality re-emphasized by the COVID-19 pandemic: that all people, systems, and generations are intrinsically interdependent and that all societies face shared risks and responsibilities. Interdependence among people emphasizes the need for equity; interdependence among systems points to the need for resilience; and interdependence across generations highlights the need for sustainability. The most important knowledge need at this time is an understanding of how to better address and harness these interdependencies” [4] (p. 8). It highlights the risk of repeating past mistakes by ignoring existing research data or not performing needed research on socio-economic recovery [4]. It is noted that socio-economic recovery needs an understanding of the interconnectedness of “health, the environment, politics, economics and societal well-being” [4] (p. 30) and it highlights the need to provide data that enables shared risk and does not unequally burden marginalized groups [4]. Many research priorities are linked to the concept of risks often in conjunction with marginalized groups such as women, workers, disabled people, poor households, etc. Some of the research priorities include

• “determining the best strategies for ensuring safe workplaces and decent work, in particular for those workers who face greater risks? [RP3.1.2]” [4] (p. 11),
• using information technologies to detect risks [RP1.1.3],
• research on the reality of global interconnectedness [RP1.4.1] and [RP1.4.2],
• generating research data to influence risk perception by involving social groups who are impacted the most by socio-economic realities [RP1.5.3] [4],
• determining how to decrease “barriers and risks associated with digital technologies while also maintaining any digital inclusion gains that occurred during the COVID-19 pandemic, such as increased opportunities for working from home, virtual healthcare visits and remote learning” [RP2.4] [4] (p. 42).

Some sub-research priorities cover

• equitable access to digital technologies [RP2.4.1],
• achieving digital and media literacy [RP2.4.4]
• high-quality education [RP2.4.3],
• how to protect children and to address gender-based and domestic violence [RP2.4.2],
• focusing on women’s realities [RP3.5.1] [RP2.5.2],
• research on new global governance structures that takes into account local realities, exhibits long term thinking and ensures the so far rare reality of the meaningful participation of marginalized groups [4],
• “to build fit-for-purpose global governance structures that enable effective responses to transnational risks while also encouraging everyone to act in solidarity in the face of shared responsibilities [RP1.4.3]” [4] (p. 30),
• how to promote mental health before during and after a disaster, especially of the ones in the greatest risk of a decrease in mental health such as marginalized groups [RP2.5],
• various sectors of the workforce at risk [RP3.1].

1.2. Risk Narrative and the UN Office for Disaster Risk Reduction (UNDRR) Strategic Framework 2022–2025

The UNDRR as the very name indicates focuses on risk [15]. In the UNDRR strategic framework 2022–2025, it is flagged that risk-informed “evidence, innovation, and good practices on risk” that also includes systemic risk has to be used to inform decision-making processes as needed [15] (p. 4). The report flags the importance of evidence-based advocacy to mobilize citizens, risk knowledge and information, and monitoring of action results [15].

“UNDRR will work with Member States and other stakeholders in development, fragile, and humanitarian contexts to improve access to enhanced products, tools, and capacity development, applicable at global, national, local, sectoral, and thematic levels responding directly to Member States expressed needs for better quality and relevant risk information. In line with Priority for Action 1 of the Sendai Framework, UNDRR, together with UN system partner agencies, will emphasize technology and data as amplifiers, including for capacity development across sectors. Faced with an increasingly complex and uncertain risk landscape, where climate change and systemic risks threaten our social, economic, and financial systems, a greater understanding of the interconnected nature of hazards, exposure, and vulnerability will be critical for effective disaster risk reduction and for achieving the Sustainable Development Goals (SDGs). Comprehensive climate and disaster risk information, innovation, and strengthening of the science-policy practice interface will be essential to achieve better risk-informed public and private decision-making and investment in building resilience” [15] (p. 10).

“Integrating human rights, gender equality and the rights of persons with disabilities into disaster risk reduction” [15] (p. 16) is a section by itself. The report lists various strategic objectives including

• the generation of quality risk information and analysis which includes systemic and cascading risk (deliverable 1.1.3),
• risk-informed and preventing future risks focused UN system policies, guidelines and inter-agency initiatives related to the Sendai Framework, and the 2030 Agenda for Sustainable Development, scientific research agenda on disaster risk reduction (Deliverable 1.1.4),
• applying to action of disaster risk reduction practices, knowledge, and methodologies (deliverable 1.2.1),
• monitoring based on data disaggregation by sex age and disability (deliverable 1.3.1),
• to develop and implement multi-stakeholder, and gender-responsive, inclusive risk governance (result 2.2 and deliverable 2.2.1) [15].

The fourth strategic objective is to “make disaster risk reduction central to sustainable development” by emphasizing advocacy and the sharing of knowledge of governments and key stakeholders [15] (p. 24).

1.3. Science and Technology Governance and Risk Narratives

Perception of risk is a social and cultural construct and is impacted by many factors [45]. Risk perception and an understanding of risk narratives of scientific and technological advancements is a critical aspect of scientific literacy [51,52,53,54,55,56]. However, risk perceptions are shaped by the values of individuals and groups [57,58,59,60]. Various studies describe the utility of artificial intelligence, machine learning, machine reasoning in relation to EDMPP, decision support systems, virtual coaches, conversational agents and robots (many cited in [16]). However, problems are noted including problems in relation to marginalized groups such as disabled people (many cited in [16]). At the same time are technologies such as artificial intelligence, machine learning, machine reasoning, Bayesian network approaches, decision support systems, virtual coaches, and conversational agents used to engage with risk narratives [61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81]. Risk narratives are mentioned extensively in conjunction with science and technology governance and various ethics discourses including environmental ethics [82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97]. A recent addition to technology governance is the effort of the National Academy of Medicine’s Committee on Emerging Science, Technology, and Innovation in health and medicine (CESTI) to generate a technology impact and governance framework [97]. In the CESTI framework it is noted that technology governance efforts have to cross sectors and disciplines and have to involve stakeholders reflecting many different groups and how benefit and risk are defined [97]. It is argued that “policies, norms, standards, and incentives of [a] particular sector” impact how a technology is developed and used, which further impacts the degree to which the benefits and risks of a technology can be maximized and mitigated, respectively [97] (p. 41). It is furthermore argued that governance efforts need to include the generation of plausible future scenarios that address potential conflicts [97] and specifically include social conflicts between individuals and groups.

1.4. Equity/Equality, Diversity, and Inclusion and Risk Narratives

Equity, diversity, and inclusion (EDI) and other EDI phrases and frameworks [46] are employed as policy concepts to improve research, education, and participation in the workplace for marginalized groups such as women, Indigenous peoples, visible/racialized minorities, disabled people, and LGBTQ2S [46] including policies in relation to environment-focused disciplines and programs [98]. Terms such as equity, equality, inclusion, and diversity are also mentioned in relation to risk narratives [99,100,101,102,103,104,105,106,107,108,109,110,111,112]. Some of the most overlooked groups in “disaster risk reduction and adaptation strategies are marginalized populations and [they] lack empowerment to utilize and take action against early warning signs (or information) of risks” [47] (p. 1). People with disabilities and their organizations are recognized by the Sendai Framework for Disaster Risk Reduction as key contributing stakeholders in disaster resilience-building at the community level [48]. Disasters and emergencies have historically had a greater impact on “racially and ethnically diverse and socioeconomically disadvantaged communities” [49,50].

Many equity, diversity, and inclusion issues were evident in the COVID-19 responses to date [113] (see for example a recent article related to disabled people and COVID-19 [114]). The UN Research Roadmap for the COVID-19 Recovery notes the following population groups as experiencing highest degree of marginalization:

“Women, older persons, adolescents, children and youth, especially girls and young women, persons with disabilities, persons with mental health conditions, Indigenous peoples, migrants, refugees, stateless and internally displaced persons, conflict-affected populations, minorities, persons in detention or in institutionalized settings (e.g., persons in psychiatric care, drug rehabilitation centres, old age homes), slum dwellers, people in informal settlements, homeless persons, persons living with HIV/AIDS and other people with pre-existing medical conditions, small farmers, fishers, pastoralists, rural workers in informal and formal markets, and other people living in remote rural areas as well as urban informal sector and self-employed who depend on market for food, the food insecure, particularly in countries affected by prolonged conflict and crisis, people in extreme poverty or facing insecure and informal work and incomes, groups that are particularly vulnerable and marginalized because laws, policies and practices do not protect them from discrimination and exclusion (e.g., LGBTI people)” [4] (p. 48).

EDI frameworks are seen to be useful tools to deal with various risks such as justice, equity, diversity, and inclusion linked to climate change risks [115]. Risks for groups covered under EDI are

• suicide of black people using Justice, Equity, Diversity, and Inclusion (JEDI) [116];
• gender and race inequalities that become conflated mentioning Athena Swan [117];
• increasing students ability to take intellectual risk using diversity, equity, and inclusion [118];
• racism against Black, Indigenous, and people of color and risk of overdose using diversity, equity, and inclusion [119];
• how transfer students have been at higher risk of attrition due to known academic and social barriers using equity, diversity and inclusion [120];
• how climate service development is relatively well resourced in places, there is a risk that the landscape becomes fragmented, duplicative, confusing, and inefficient using equality, diversity and inclusion, for example for gender, under-represented regions, and under-represented stakeholder groups [121];
• mitigating culture risk through diversity, equity, and inclusion efforts [122];
• how diversity, equity and inclusion efforts within their institutions will drive innovation and reduce risks [123];
• how diversity, equity and inclusion can improve risk related issues mentioned in the Global Assessment Report on Disaster Risk Reduction [124];
• how medical trainees encounter risks of stigma are seen as a vulnerable group and equity, diversity, and inclusion is needed [125];
• diversity, equity, and inclusion in medical regulation [126];
• diversity, equity, and inclusion for undergraduate students and risks associated with ecological fieldwork [127].

It is also noted that decreasing EDI efforts carries risks [128]. For example, there are many problems that have come in the aftermath of COVID-19 for disabled people [129] and other marginalized groups. These problems might hinder diversity, equity, and inclusion with a specific risk that the recent gains in equity and advancement for women may be lost [130]. Similarly, there is a risk of simply learning to exercise cognitive-emotional skills, such as perspective-taking with those who are similar to oneself without diversity, equity, and inclusion [131]. By not employing equality, diversity and inclusion in geomorphology is creates danger of reputational risk for geomorphological groups and organizations “like the British Society for Geomorphology, and ensure[s] that the many potential benefits of geomorphology for science and society remain underutilized at best” [132] (p. 5). Championing for diversity, equity, and inclusion comes with a personal and professional risk [133] (see also [134,135]), and risk related organizations such as the Risk Management Association have EDI statements [136].

The National Academy of Medicine’s Committee on Emerging Science, Technology, and Innovation in health and medicine (CESTI) proposed technology impact and governance framework [97] includes principles that are also used within EDI discussions such as justice. In the CESTI framework justice is described as “equity between groups faced with structural and systemic inequalities, a fair distribution of risks and benefits of technologies” [97] (p. 44). The CESTI framework uses other principles relevant to EDI and risk narratives such as fairness, autonomy/individual and group self-determination, and collective good with concepts such as solidarity, civic responsibility and stewardship, and individual good whereby risk benefit is linked to the principle of justice, collective, and individual goods [97] (figure page 45).

2. Materials and Methods

2.1. Study Design and Research Questions

Scoping studies are useful in identifying the extent of research that has been conducted on a given topic [137,138]. The study asked three questions: (1) How often are the terms “risk” and “social risk” mentioned in the literature covered (2) How often and how are the terms risk perception, risk governance, risk literacy, risk communication, risk education and risk narrative mentioned in the literature covered (3) How often and how is the term risk and the risk related concepts risk perception, risk governance, risk literacy, risk communication, risk education and risk narrative engaged with in the EDMPP literature in conjunction with EDI groups. The study employed a modified version of a scoping review outlined by Arksey and O’Malley [139] as described in [140].

2.2. Data Sources, Data Collection (Search Strategies) and Inclusion Criteria

On 14 April 2022, the academic databases EBSCO-HOST (an umbrella database that includes over seventy other databases itself), SCOPUS (which incorporates the full Medline database collection), the databases accessible through Compendex, which include IEEE sources and the Web of Science database were searched with no time restriction. These databases were chosen as they contain relevant content covering risk including in relation to EDMPP and technologies and governance of both. Inclusion criteria were as described before in [140]. The following search terms and strategies were used (

Table 1. Search strategies used to obtain abstracts (first search term) for online manifest coding of terms related to risk and abstracts downloaded with specific topics (second search term) for desktop hitcounts manifest coding and qualitative thematic analysis of the term risk used in conjunction with EDI groups.

Strategy	Sources Used	First Search Term (Abstract)	Second Search Term (Abstract)
Search strategies for manifest coding of online databases
Strategy 1	SCOPUS/EBSCO-HOST/Compendex/Web of Science	(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness”)
Strategy 2	SCOPUS/EBSCO-HOST/Compendex/Web of Science	COVID	-
Strategy 3	SCOPUS/EBSCO-HOST/Compendex/Web of Science	Emergency
Strategy 4	SCOPUS/EBSCO-HOST/Compendex/Web of Science	Disaster
Strategy 5	SCOPUS/EBSCO-HOST/Compendex/Web of Science	(“e-coaching”) OR (“decision support system”)	-
Strategy 6	SCOPUS/EBSCO-HOST/Compendex/Web of Science	Bayesian network	-
Strategy 7	SCOPUS/EBSCO-HOST/Compendex//Web of Science	Conversational agent
Search strategies for obtaining abstracts for desktop manifest and thematic coding
Strategy 8	SCOPUS/EBSCO-HOST/Compendex/Web of Science	“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness”	“artificial intelligence” OR “machine learning” OR “robot*” OR “quantum*” OR “machine reasoning”
Strategy 9	SCOPUS/EBSCO-HOST/Compendex/Web of Science	“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” “disaster preparedness” OR “emergency preparedness”	“disabl*” OR “disabili*” OR “impairm*” OR “deaf” OR “neurodiver*” OR “autism” OR “adhd” OR “impair*”
Strategy 10	SCOPUS/EBSCO-HOST/Compendex/Web of Science	(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” “disaster preparedness” OR “emergency preparedness”)	“patients”
Strategy 11	SCOPUS/EBSCO-HOST/Compendex/Web of Science	(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness”)	COVID
Strategy 12	SCOPUS/EBSCO-HOST/Compendex/Web of Science	Bayesian network*	COVID
Strategy 13	SCOPUS/EBSCO-HOST/Compendex/Web of Science	Bayesian network*	emergency OR disaster
Strategy 14	SCOPUS/EBSCO-HOST/Compendex/Web of Science	“conversational agent*”	(emergency OR disaster OR COVID)
Strategy 15	SCOPUS/EBSCO-HOST/Compendex/Web of Science	(“e-coaching”) OR (“decision support system”)	(equality OR diversity OR inclusion OR equity)
Strategy 16	SCOPUS/EBSCO-HOST/Compendex/Web of Science	(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness” OR “emergency)	(equality OR diversity OR inclusion or equity)
Strategy 17	SCOPUS/EBSCO-HOST/Compendex/Web of Science	COVID	Social Risk
Strategy 18	SCOPUS/EBSCO-HOST/Compendex/Web of Science	Disaster*	Social risk
Strategy 19	SCOPUS/EBSCO-HOST/Compendex/Web of Science	Emergenc*	Social risk

).

2.3. Data Analysis

To answer the research questions, we performed two types of analysis. We performed a descriptive quantitative analysis approach [141,142] (manifest coding [143,144]) providing frequencies of search terms using (a) online search engines of the databases for the keyword search strategies 1–7 (Table 1, Figure 1) and (b) 11,675 downloaded abstracts obtained through strategies 8–19 (Table 1 and Figure 2). The manifest coding of the abstracts was performed as described in [140]. In short duplicates of abstracts were eliminated using the software Endnote and the abstracts obtained with each of the strategies 8–19 were made into PDF’s and the abstracts in the various PDF documents were searched with the ‘CTRL F’ function of Adobe Acrobat software. We also performed a thematic analysis for the abstracts downloaded from strategies 8–19 using the qualitative software analysis tool NVIVO 12 to identify how risk and “social risk” were mentioned in relation to EDI groups including disabled people, LGBTQ2S+, Indigenous people, ethnic groups, and women. Manifest coding and qualitative thematic analysis were performed by both authors and differences were resolved through peer debriefing and the audit trail for the thematic analysis was present in the use of codes and comments and memos in the NVIVO 12 software.

3. Results

3.1. Quantitative Hit Count Results

3.1.1. Manifest Coding for the Terms “Risk*” and “Social Risk*”

Table 2. Manifest coding for the terms “risk*” and “social risk*”.

Mentioning in the 3,404,903 Online Abstracts
Term 1	Abstracts in Online Databases	Risk	Social Risk
COVID Online search	847,850	162,382	96 abstracts in NVivo
Emergency Online search	1,847,895	270,882	447 abstracts in NVivo
Disaster Online search	457,339	92,816	388 in NVivo
“Bayesian network*” Online search	69,987	12,002	14
(“e-coaching”) OR (“decision support system”) Online search	100,036	10,962	10
(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness”) Online search	75,243	29,121	33
“Conversational agent*” Online search	6553	182	0
Mentioning in the 11,675 abstracts downloaded to desktop
Term 1	Abstracts on computer	Risk	Social risk
(ABS (“risk*”) AND ABS (“emergency” OR “COVID” OR “disaster”) AND ABS (“people with disabilities” OR “disabled people”))	506 abstracts in NVivo	2492 hits, not abstracts	657 hits, not abstracts
e-coaching alone	228 abstracts on computer in NVivo	17	0
“disaster management” OR “emergency management” OR “emergency planning”OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness”	“artificial intelligence” OR “machine learning” OR “robot*” OR “quantum*” OR “machine reasoning” 656 abstracts in NVivo	20	0
“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” “disaster preparedness” OR “emergency preparedness”	“disabl*” OR “disabili*” OR “impairm*” OR “deaf” OR “neurodiver*” OR “autism” OR “adhd” OR “impair*” 529 abstracts in NVivo	352 hits, not abstracts mostly medical risk	0
(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” “disaster preparedness” OR “emergency preparedness”}	“patients” 2486 abstracts in NVivo	738 hits, not abstracts mostly medical risk	0
(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness”}	COVID 885 abstracts in NVivo	607 hits, not abstract	1
“Bayesian belief network*”	2091 in NVivo	555/681/365/288	0
“Bayesian network*” AND COVID	74abstracts in NVivo	145 hits, not abstract	0
“Bayesian network*” AND (emergency OR disaster)	755 abstracts in NVivo	1156 hits, not abstracts	1
(“e-coaching”) OR (“decision support system”) AND (equality OR diversity OR inclusion OR equity)	800 abstracts in NVivo	307 hits, not abstracts	1
(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness” OR “emergency AND (equality OR diversity OR inclusion or equity)	797 abstracts in NVivo	700 hits, not abstract	1
“Automated decision support”	420 abstracts in NVivo	195 hits, not abstract	0
e-mentoring	355 abstracts in NVivo	8	0
(“virtual coaching” OR “virtual coach*”)	319 abstracts in NVivo	14	0
“Automated Dialogue”	141 abstracts in NVivo	7	0
“conversational agent*” AND (emergency OR disaster OR COVID)	231 abstracts in NVivo	9	0

shows that “social risk” is rarely mentioned especially in conjunction with files that focused on technologies.

3.1.2. Manifest Coding for the Terms “Risk Governance”, ”Risk Education”, “Risk Narrative” and “Risk Perception”

Table 3. Manifest coding for the terms “risk governance”, ”risk education”, “risk narrative*”, “risk literac*”, and “risk perception*”.

Mentioning in the 3,404,903 Online Abstracts
Term 1	Abstracts in Online Databases	“Risk Narrative*”	“Risk Perception*”	“Risk Governance”	“Risk Education”	“Risk Literac*”	“Risk Communication*”
COVID Online search	847,850	3	3572	57	6	5	1843
Emergenc * Online search	1,847,895	1	1461	112	49	7	2351
Disaster * Online search	457,339	3	2110	584	85	16	1688
“Bayesian network *” Online search	69,987	0	29	1	0	0	17
(“e-coaching”) OR (“decision support system”) Online search	100,036	0	27	3	0	0	22
(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness”) Online search	75,243	1	595	101	14	8	859
“Conversational agent *” Online search	6553	0	1	0	0	0	7
Mentioning in the 11,675 abstracts downloaded to desktop
Term 1	Abstracts on computer	“Risk narrative*”	“Risk perception*”	“Risk Governance”	“Risk Education”	“Risk Literac*”	“Risk communication*”
COVID+ “social risk”	96 abstracts in NVivo	0	24	1	0	0	2
Emergency + “social risk”	447 abstracts in NVivo	0	44	6	0	0	10
Disaster + “social risk”	388 abstracts in NVivo	0	49	9	0	0	15
(ABS (“risk*”) AND ABS (“emergency” OR “COVID” OR “disaster”) AND ABS (“people with disabilities” OR “disabled people”))	506 abstracts in NVivo	0	41	4	0	0	13
e-coaching alone	228 abstracts on computer in NVivo	0	-	0	0	0	0
“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness”	“artificial intelligence” OR “machine learning” OR “robot*” OR “quantum*” OR “machine reasoning” 656 abstracts in NVivo	0	1	0	0	0	0
“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” “disaster preparedness” OR “emergency preparedness”	“disabl*” OR “disabili*” OR “impairm*” OR “deaf” OR “neurodiver*” OR “autism” OR “adhd” OR “impair*” 529 abstracts in NVivo	0	5	0	0	0	0
(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” “disaster preparedness” OR “emergency preparedness”)	“patients” 2486 abstracts in NVivo	0	6	0	0	0	4
(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness”)	COVID 885 abstracts in NVivo	0	19	3	0	0	24
“Bayesian belief network*”	2091 in NVivo	0	6	0	0	0	1
“Bayesian network*” AND COVID	74abstracts in NVivo	0	1	0	0	0	0
“Bayesian network*” AND (emergency OR disaster)	755 abstracts in NVivo	0	0	0	0	0	0
(“e-coaching”) OR (“decision support system”) AND (equality OR diversity OR inclusion OR equity)	800 abstracts in NVivo	0	1	2	0	0	0
(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness” OR “emergency AND (equality OR diversity OR inclusion or equity)	797 abstracts in NVivo	0	7	14	0	0	7
“Automated decision support”	420 abstracts in NVivo	0	1	0	0	0	0
“e-mentoring”	355 abstracts in NVivo	0	0	0	0	0	0
(“virtual coaching” OR “virtual coach*”)	319 abstracts in NVivo	0	0	0	0	0	0
“Automated Dialogue”	141 abstracts in NVivo	0	0	0	0	0	0
“conversational agent*” AND (emergency OR disaster OR COVID)	231 abstracts in NVivo	0	0	0	0	0	1

shows very little to no engagement with most risk concepts we covered.

3.1.3. Risk and EDI Groups

Table 4. Frequency of EDI groups mentioned in conjunction with the term “risk*” in the 11675 abstracts (not number of abstracts).

Term 1	Term 2	Disabled or Disabili*	Deaf	“Mental Health”	Autism or ADHD or “Attention Deficit” or Neurodiverse	Impair*	Women or Gender or Girls	“Indigenous Peoples” or Aboriginal or “First Nations” or Metis or Inuit	LGB*	“Visible Minorit*” or “Racialized Minorit*” or Ethnic* or Black
COVID+ “social risk” 96 abstracts in NVivo	-	0	0	17	0	1	12	2	0	10
Emergency + “social risk” 447 abstracts in NVivo	-	4	1	7	0	3	27	1	0	7
Disaster + “social risk” 388 in NVivo	-	4	1	3	0	3	10	1	0	1
(ABS (“risk*”) AND ABS (“emergency” OR “COVID” OR “disaster”) AND ABS (“people with disabilities” OR “disabled people”) 506 abstracts in NVivo	Risk*	174	5	6	0	7	21	2	0	1
e-coaching alone228 abstracts on computer in NVivo	Risk*	0	0	1	0	1	2	0	0	0
“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness” “artificial intelligence” OR “machine learning” OR “robot*” OR “quantum*” OR “machine reasoning” 656 abstracts in NVivo	Risk*	1	0	0	0	0	0	0	0	0
“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” “disaster preparedness” OR “emergency preparedness” “disable*” OR “disabili*” OR “impairm*” OR “deaf” OR “neurodiver*” OR “autism” OR “adhd” OR “impair*” 529 abstracts in NVivo	Risk*	101	3	8	0	26	10	1	0	1
(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” “disaster preparedness” OR “emergency preparedness”) AND “patients” 2486 abstracts in NVivo	Risk*	3	0	13	0	7	15	0	0	0
(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness”) And COVID 885 abstracts in NVivo	Risk*	10	0	15	0	1	3	0	1	2
“Bayesian belief network*” 2091 in NVivo	Risk*	0	0	0	0	3	4	0	0	0
“Bayesian network*” AND COVID 74 abstracts in NVivo	Risk*	0	0	1	0	0	0	0	0	0
“Bayesian network*” AND (emergency OR disaster) 755 abstracts in NVivo	Risk*	0	0	0	4	0	1	0	0	0
(“e-coaching”) OR (“decision support system”) AND (equality OR diversity OR inclusion OR equity) 800 abstracts in NVivo	Risk*	1	0	0	0	3	2	0	0	2
(“disaster management” OR “emergency management” OR “emergency planning” OR “disaster planning” OR “disaster preparedness” OR “emergency preparedness” OR “emergency AND (equality OR diversity OR inclusion or equity) 797 abstracts in NVivo	Risk*	13	0	6	0	1	31	0	0	2
“Automated decision support” 420 abstracts in NVivo	Risk*	0	0	0	0	0	0	0	0	0
“e-mentoring” 355 abstracts in NVivo	Risk*	1	0	0	0	0	0	0	0	0
(“virtual coaching” OR “virtual coach*”) 319 abstracts in NVivo	Risk*	0	0	0	0	1	0	0	0	0
“Automated Dialogue” 141 abstracts in NVivo	Risk*	0	0	0	0	0	0	0	0	0
“conversational agent*” AND (emergency OR disaster OR COVID) 231 abstracts in NVivo	Risk*	0	0	0	0	0	0	0	0	1

shows a very low to no engagement with risk in conjunction with EDI groups.

3.1.4. Risk and Science and Technology Governance Terms and Ethics Fields

We found no hits for “democratizing science and technology” OR “participatory technology assessment” OR “technology assessment” OR “parliamentary technology assessment” OR “anticipatory governance” OR “upstream engagement” OR “responsible innovation” OR “responsible research and innovation” OR “transformative vision assessment” OR “AI-ethics” OR “bioethics” OR “computer science ethics” OR “information technology ethics” OR “nanoethics” OR “neuroethics” OR “robo-ethics” in the 11,675 abstracts in conjunction with the term “risk*”.

3.2. Qualitative Analysis

3.2.1. Risk Related to EDI Groups

Women, Gender, and Girls

Fifteen abstracts mention the terms women, woman, girl* or gender in conjunction with the term risk. Four were on the topic of breast cancer [145,146,147,148] engaging with an automated decision support system [145], a Bayesian Classifier [146], a Bayesian belief network [147], and shared decision-coaching, between nurses-led decision coaching, and an evidence-based decision aid [148] for evaluating risk. Another disease covered was osteoporosis where the focus was also on a decision support system using data mining ensemble technology [149]. Women are mentioned in relation to risk perception [150,151] and indicates that there is a gender difference in risk appetite [152].

As to COVID-19, risk is mentioned in relation to preparatory purchasing of safety and health care products [153] and risk factors for adverse outcomes among pregnant and postpartum women with acute respiratory distress syndrome [154].

One abstract mentioned socio-economic impacts in disasters such as the 2011 earthquake and tsunami in Japan covering Thai women [155]. Another explored the psychosocial effects of disaster among pregnant and postpartum women aged 18–45 Post-Hurricane Isaac [156] and in two abstracts it is argued that women experiencing disasters are at higher risk of mental health problems [157,158].

In one abstract it is noted that “women and girls are disproportionately exposed to risk during and in the aftermath of disaster” that there “physical health and sanitation risks unique to girls and women, including a lack of appropriate resources and private facilities” and that “despite great progress toward gender equity in Nepal in recent decades, pre-existing risk factors and embedded gender beliefs intersected with novel disaster-induced stressors to produce a range of health and security risks for women and girls. Incorporation of existing frameworks for gender-mainstreaming in disaster preparedness and response efforts is thus crucial to improve inclusivity in risk reduction“ [159] (p. 102622).

One abstract focused on the role of women in disaster risk governance and outlined “socio-cultural factors, individual characteristics, legal and institutional factors and socio-economical factors as the barriers” [109] (p. 1187). One flagged the lack of coverage of the impact of gender discrimination in disasters including disaster risk management [112].

Risk Narrative and Disabled People

After reading the abstracts containing disabled people using the search terms “disabled” OR “disabili” OR “deaf” OR “mental health” OR “autism” OR “adhd” OR “attention deficit” OR “neurodiv*” OR “impair*” generated and risk (Table 3), seventy-four abstracts were identified as relevant.

COVID-19 is covered in relation to emergency preparedness and response (EPR) and disability-inclusive disaster risk reduction (DiDRR) [160], risk of severe illness [161,162,163,164,165,166,167], increased risk of contracting COVID-19 [166,167,168,169,170,171,172] and other problems [173,174], increased mortality [175], increased risk of exclusion from treatment [176], risk of residential settings [177], and that disabled people are ignored as a high risk group [165,178]. Disabled people are at an increased risk of poverty and it is likely this risk will be increased further by COVID-19 [179]. Disabled people are also largely absent from risk-related discussions [169] and are at higher risk of lower psychosocial well-being [180,181], increased risk for abuse [182], and risk of being neglected and being subjected to restrictive measures [183]. Various abstracts highlighted that COVID-19 made the existing social problems for disabled people worse.

“While Coronavirus Disease 2019 (COVID-19) does not discriminate against particular groups, our social structures and systems mean some people are more at risk in a pandemic context—from both the disease and the social and policy responses to the pandemic. This is particularly so for people with disability, in part because they often have poorer health outcomes from underlying conditions but also due to discrimination and social exclusion. Here, we draw from a survey about the impacts of the COVID-19 pandemic on Australian children and young people with disability and their families. Respondents faced a range of inequities prior to the pandemic, and COVID-19 has further exposed and often exacerbated them. We conclude that recent developments in the Australian disability context to personalize services have arguably made people with disability and their families less safe within a pandemic context, and we outline some ways in which these issues might be addressed” [184] (p. 1192), see also for Australia [185] and Japan [186].

Covering EDMPP beyond COVID-19, various abstracts mentioned that disabled people are at higher risk [187,188,189,190,191,192,193,194,195,196,197,198,199] and that there is a lack of consideration for disabled people [47,192,200,201,202,203]. Further, there is a lack of inclusivity in involvement [202,204,205,206,207,208], and disabled people should not be forgotten in risk strategies [209]. Risk factors for disaster and emergency vulnerability among disabled people identified were a “lack of personal disaster preparedness, [and] public information that is not disability accessible and social isolation” [210] (p. 190), see also [193]. There is a lack of community preparedness and insufficient structures to assist disabled people in the disaster response or recovery phases [193,211,212], lack of resources [198], algorithmic health discrimination [213], food insecurity [214], and lack of disability inclusive disaster education [193]. It is mentioned that children with disabilities are at a greater risk of separation from their families in emergencies, unable to escape [215]. In one abstract it was flagged that “15% reported using electrically powered medical devices that might be at risk in a power outage” [216] (p. S621).

The Sendai Framework for Disaster Risk Reduction is mentioned [160,203,217,218,219] as a positive example if employed [220] and that disaster risk recovery should be used as an opportunity to fix the systemic problems disabled people faced before the disaster, ensuring not to exacerbate the old systemic problem [210]. However, it is noted that “the state of the art of emergency preparedness and disaster risk reduction for people with disabilities is at a low level” [221] (p. 37), although, it could be improved quickly if existing innovations [221] and technologies [222] such as twitter [223] were to be applied correctly. It is noted “that slight variations on the definitions of disaster or disability may increase marginalization” [217] (p. 155). Currently, Australia has no national standard for disaster risk reduction in relation to disabled people [224]. Disaster risk reduction strategies have to increase the autonomy of disabled people [210] and empower them [225]. It is also argued that “(1) intersectionality of disability with other dimensions of vulnerability warrants focused consideration; (2) enhanced disaster preparedness requires more attention in order to empower people with disabilities; and (3) negative cultural attitudes need to shift to enable purposeful inclusion of people with disabilities” [217] (p. 155). The United Nations (UN) Convention on the Rights of People with Disabilities is recognized as demanding inclusion of disabled people in disaster risk reduction [219]. Two abstracts argued for the used of the capability approach in disaster risk reduction [226,227] and one for the need for social and environmental protection embedded in disaster risk mitigation [189]. Risk communication is discussed [228] and the Washington Group Short Set of Questions on Disability (WGQ) has been promoted to support inclusive practice within Disaster Risk Reduction (DRR), humanitarian action, and more broadly international development [229]. Risk of injury [205,230], disability [188,231], and “architectonical barriers and cultural ones make people with disabilities the most vulnerable population” [230] (p. 355), ee also [201]. One asks “what are the current shortfalls (physical and cultural) that put people with disabilities at risk during emergencies and what can be done to improve these” [222] (p. 1).

In one abstract it is argued “that tales of communities in crisis commonly depoliticize disaster. By inscribing the disabled body with a narrative of “natural” vulnerabilities and inevitable suffering, conventional disaster discourse obscures the political significance of structural inequalities that render people with disabilities more at risk in disaster” [212] (p. 51). It is further argued that “Disability inclusive disaster risk reduction strategies need to focus on strengthening key environmental and personal factors that have a fluctuating influence on each individual’s ability to manage their wellbeing at the different time points over the course of a cyclone event” [232] (p. 1). In one abstract it is argued that “disability-related terms and concepts such as accessibility, inclusion, and universal design” are applicable to everyone [233] (p. 140)

One abstract focused on women with disability, stating that “Women with disabilities are disproportionately impacted by disasters. • Exclusionary social processes cause heightened risk exposure and reduce resilience. • Inclusive disaster risk reduction must deliver support to those most at risk. • Disaster risk reduction practices can contribute to greater social equity. Despite increased international commitment to disability-inclusive disaster risk reduction (DiDRR) people with disabilities remain largely unseen, unheard, and unaccounted for in DRR processes and planning. This is most marked amongst women with disabilities who experience specific gender, disability, and poverty-based disadvantages, which disasters exacerbate. Our research found that women with disabilities are disproportionally impacted by disasters and are the least able to access institutional support across the preparedness, response, and recovery phases of disaster events. Furthermore, the increased threat of violence following disasters heightens their risk of additional harm. In the absence of formal supports women with disabilities have few choices but to rely upon the social capital of their households and neighbours for assistance. They ‘recover’ in whatever ways they can—through short-term loans, reduced food consumption and/or migration—each carry significant costs to their longer-term resilience. This paper unpacks the root causes of women with disabilities’ marginalisation in disaster contexts, many of which are extensions of exclusionary processes that play out in their daily lives. We also present steps to position women at the centre of DRR discourse, which will benefit all” [234] (p. 1).

LGBTQ2S+

Covering LGBTQ2S+ using search terms such as “gay” or “lesbian” or “LGB*” one abstract stated “International emergency management and disaster risk reduction policies and planning have rarely included lesbian, gay, bisexual, transgender, intersex, and queer (LGBTIQ) people’s specific health and wellbeing concerns, despite increasing research showing that these groups face some specific vulnerabilities and additional challenges” [235] (p. 647).

Indigenous People

Using the search terms “Indigenous People*” OR “Aboriginal” OR “First Nations” OR “Inuit”, OR “Metis” one abstract noted that indigenous people were at a higher risk of exposure and susceptibility to infection or complications because of high poverty rates and associated social risks such as “homes without indoor plumbing or access to running water, which precludes effective hand hygiene measures and promotes disease acquisition and spread” [236] (p. 207).

Ethnic

Eight abstracts covered ethnic groups using search terms “ethnic” OR “racialized” OR “minort*”.

One abstract flagged the undercounting of ethnic/racial populations to be at risk of floods (by using standard census tools) and that environmental justice impacts of flood risks is hard to evaluate with the tools used, all of which hinders EDMPP [237]. One study found measurable inequity in exposure to flood risk across age groups, education level, and income status but not ethnic/racial groups [238]. One study focused on racial/ethnic and socioeconomic characteristics of populations living flood risk zones [239].

One study covering COVID-19 concluded that “efforts may also be needed to reduce structural racism and address social risk factors to improve quality of care and population health in communities of color” [240] (p. 1345). One study indicated “that people who self identify as mestizo ethnic have clinical risk factors of “arterial Hypertension, Diabetes Mellitus type 2, Bronchial Asthma and immunosuppression status” and social risk factors of “overcrowding and living alone” [241] (p. 159).

Finally in one abstract it is stated “Emerging evidence has indicated a negative and disproportionate impact of COVID-19 on Black, Asian, and minority ethnic (BAME) communities. Previous studies have already reported that biological and social risk factors increase disease susceptibility, particularly in BAME communities. Despite frontline workers in ethnic minority communities, the United Kingdom’s National Health Service is attempting to quell the pandemic, disproportionate numbers of BAME physicians and other health care workers have died of COVID-19” [242] (page 1).

3.2.2. Social Risk Related to EDI Groups

Women, Women, Girl*, and Gender

Five abstracts were found. One study found covered the social determinants of health, namely “housing problems, behavioral health issues, disability, and neighborhood-level stress” where the demographics of study subjects were listed as “57% were women, 25% dually diagnosed, 67% White/non-Hispanic, 18% unstably housed, and 37% disabled”. This suggests that “to set appropriate benchmarks for comparing health plans, quality measures for emergency department visits should be adjusted for both medical and social risks” [243] (p. 362). In one abstract it is stated that oxytocin increased acceptance of social risk in men and women [244]. In another abstract it is argued that “enhancing the legitimacy of women’s compensation requests does not eliminate the social risk of asking, and that eliminating the social risk of asking is not sufficient to legitimize their requests” [245] (p.1). One study looked at the context of the social risk of “the public model of long-term care with cultural values that place migrant women in these activities” and flagged that “the contradictions that exist in the welfare system: care workers suddenly become essential while they are excluded from access to basic social rights” [246] (p. 1). Lastly, one study stated that “clinical and social risk factors and barriers to access health care are associated with adverse outcomes among maternal cases of COVID-19 ARDS in Brazil” [154] (p. 415).

Disabled People

Using the search terms “disabled” OR “disabili*” OR “deaf” OR “mental health” OR “autism” OR “adhd” OR “attention deficit” OR “neurodiv*” OR “impair*”, only abstracts covering mental health were found. Three abstracts indicated social risks in relation to mental health all of which were related to COVID-19 [247,248,249]. Two of them mentioned women and girls [248,249]. One stated “how the lockdown measures affected the general population’s mental health in Colombia and highlight some social risk factors in health” [248] (p.1). One abstract mentioned “social risk factors and pandemic stressors will contribute to negative mental health outcomes, especially among vulnerable populations” [247] (p. 5297). Adolescent girls and young women (AGYW) in South Africa face various psycho-social risk factors which already disproportionally affect the mental health of AGYW in these communities; the COVID-19 pandemic intersects with these pre-existing social and environmental factors [249].

Two were related to mental health but not COVID-19. One highlighted the need “to build a service program and individual client capacity to improve mental health-related quality of life among individuals at risk for depression, with exposure to social risk factors or concerns about environmental hazards in areas of Southern Louisiana at risk for events such as hurricanes and storms” [250] (p. 1683). Another abstract covered “indices of social risk among first attenders of an emergency mental health service in post-conflict East Timor” [251] (p. 929). One stated that sick building syndrome symptoms causes psycho-social risk factors such as anxiety and stress as well as ergonomic risk factors [252].

Indigenous Peoples, Ethnic Groups and LGBTQ2S+

Six abstracts were found for ethnic groups, none of which covered Indigenous people or members of the LGBTQ2S community.

In one abstract it is stated that “social risk factors increase disease susceptibility, particularly in BAME communities” [242] (p. e22381). One study looked at “mental health symptoms and social risks during COVID-19, compared to before the pandemic, for urban, racial and ethnic minority school-age children” [253] (p. 1753). In one abstract it is argued that “accounting for Social Risk Does not Eliminate Race/Ethnic Disparities in COVID-19 Infection” [254] (p. 1183). One study covering COVID-19 concluded that “efforts may also be needed to reduce structural racism and address social risk factors to improve quality of care and population health in communities of color” [240] (p. 1345). In one abstract it is argued that social risks of EDMPP become more problematic and four types of social risk are identified “for the Chinese minority regions”. These include “the livelihood risk, the stability risk, the identity risk and the governance risk” [255] (p. 700). It is also argued in the abstract that to “minimize and eliminate the social risk and social crisis in the minority regions, we should take their ethnic and regional features into consideration” [255] (p. 700). In one abstract it is argued that operators from outside have to understand “Latin America’s unique administrative framework, property rights, water, infrastructure, Indigenous groups, conflicts, and insecurity and workforce” to better understand “potential social risks and impacts” [256] (p. 1414).

3.2.3. EDMP and “Risk Governance”, “Risk Communication”,” Risk Education”, “Risk Narrative” and “Risk Perception”

Risk education, risk literacy and risk narrative generated no hits in the downloaded abstract.

Risk Governance

In one abstract risk governance is defined as “concept of risk governance pertains to the many ways in which multiple actors, individuals, and institutions, public and private, deal with risks. It includes formal institutions and regimes and informal arrangements” [257] (p. 434) and seven criteria for risk governance are outlined: “(a) life cycle thinking, (b) triple bottom line, (c) inclusion of stakeholders, (d) risk management, (e) benefit–risk assessment, (f) consideration of uncertainty, and (g) adaptive response” [61] (p. 1). Applications for risk governance mentioned were urban challenges [257], nanotechnology and sustainability [61], national disaster [258], big data security [259], landslides [260], public security related to big data [261], communicable diseases and natural disasters [262], disaster management frameworks [263], disaster risk reduction measures [109], GSI adoption [264] and infrastructure [264]. Barriers for risk governance mentioned were “lack of knowledge about risk assessment and emergency planning” and “divergent, sometimes even opposite, stakeholders’ views on several issues” [260] (p. 27). One abstract mentioned the vulnerability analysis matrix as a risk governance tool [265]. As to EDI related groups women were mentioned [109] as part of the need for multi stakeholder involvement [109], stakeholders and many different voices [260]. As to “social risk governance” one abstracts made the case for the usefulness of big data technology for social risk governance [266]. Another the use of “framework Adaptive Control Theory” [267] and another “social network analysis based agent-based modeling” [264].

Risk Perception

Risk perception was mentioned in conjunction with water scarcity [268], COVID-19 related health [269], fire [270,271], hurricane storm tide [272],terrorism [273], disaster [187,274,275,276], environmental problems [277], online shopping [278], water security [279], genetically modified food [280], volcanic risk [281], crimes, economic risks, uncontrollable risks, accidents, environmental risks, natural disasters, and future risks [282], public emergencies [283], flood [284] check-in services [285], public safety [286] social security and financial security [287], petition [288], radioactive waste [151], nuclear power [151], flash flood [289], urban context [187] natural hazard [275], severe weather [290], abrupt geological hazards in coastal rural area [291], emergence and spread of COVID [292], natural disaster [293] disease [270], public health emergencies [294] green stormwater infrastructure [295] earthquake [275] and typhoon [275].

As to making visible and improve risk perceptions including social risk perception, the following technologies were mentioned: visualization tools [272,281,296], social media [292,297], agent-based simulation [298], cross-modal semantic fusion, temporal knowledge graph and analysis, complex social network intelligent decision-making methods [287], autoregressive integrated moving average (ARIMAbased time-series prediction model [299], structural equation (SEM) model [300], fuzzy cognitive maps [301], unsupervised and supervised machine learning algorithms, Bayesian Belief Networks [295] and decision support system [271]. Technological risk perception is defined in one abstract as the “processing of physical signals and/or information about a potentially harmful impact of using technology and the formation of a judgment about seriousness, likelihood, and acceptability of the respective technology” [302] (p. 293).

Groups mentioned in conjunction with risk perception were: farmers [268,303], immigrants [293], Canadian born adults [293] the public [150,270,294,295,299], citizens [277,283,289,304], locals [284], Koreans [282], community [281], household with disabled people [275], people with disabilities [187], consumers [278,280,305], water professionals [279] undergraduate college students [290], mobility-disadvantage group [291], public agencies [292], hospital authorities [276], people engaged in nuclear business [150,151], women [151] and gender [280].

As to EDI related groups from the group mentioned, household with disabled people [275], people with disabilities [187], mobility-disadvantage group [291], women [151] and gender [280].

One abstract found that “households with disabled members are less likely to prepare emergency kits and to plan evacuation. However, with the adjustment of risk perception (probability, consequence, worrisome) and other factors experience of earthquake and typhoon hazards, home ownership status, whether there are children in the home, perceived social status, family income, gender, age, education attainment, and religious status the differences in adopting all 6 preparedness activities between households with disabled members and households without disabled members become nonsignificant” [275] (p. 575). One stated “Some of consumer characteristics including gender, education background, personal annual income, and with at least one child under 18, significantly influenced their risk perception of GMF [genetically modified food]” [280] (p. 30).

Risk Communication

Risk communication was linked to health risk related to public health [292], “individual hurricane evacuation intentions during the COVID-19 pandemic [306] (p. 507), disaster risk reduction [307], prevention and control of coronavirus disease (COVID-19) [308], COVID-19 [308,309,310,311,312,313,314,315,316,317], emergency planning and response [318], cyclone Amphan [319], health communication [312], for sustainable society [320], terrorism [273], public health information [321], household flooding [322], climate change [323], Sustainable Disaster Risk Reduction [324], volcanic crisis [281], earthquake [325], nuclear issues [150], 2008 earthquake in Sichuan province in China [227] and radioactive waste [326]. To give a few quotes: “risk communication for shaping responses and policies in an efficient and coordinating way” [261] (p. 1), “prevention and mitigatory role of risk communication” [309] (p. 1), “vital to creating the required awareness and preparation” [273] (p. 1), “risk communication should emphasize the importance of both resource- and action-based preparedness” [327] (p. e206881) and “utility of risk communication techniques designed to reduce psychological distance” [323] (p. 957).

“Risk communication is necessary to solve complicated social problem between stakeholder with different knowledge and different standard about risk” [328] (p. 310) and “The need for fair risk communication has emerged as a result of a more global and more flexible economy as well as of a media dominated world” [329] (p. 1007)

For and with whom the following groups are mentioned: stakeholders [307,310,328,330,331], African American Churches [332], community [314,333], public health organizations [318], acute hospitals [318], vulnerable populations [319], many different groups of society taking into account cultural, social, political and economic factors [312], medical residents [316,322], differing segments of the population [334], nuclear experts [150], women in metropolitan areas [150], people with disabilities [228], the public [326,329], businesses [329], Bangladeshi adults [317], tourists [324], public health authorities [331], local governments [331], media [331], risk informed communities [281] and local communities [325].

Finally, in one abstract it is stated: “Research on risk communication audiences advanced on risk perception and multiway engagement with notable interest in personal factors such as gender, race, age, and political orientation” [335] (p. 2240).

As to risk communication problems, the following are mentioned: constantly challenged and complex and difficult [307], media framing [319] “low preparedness [273] and it is recognized that help with risk communication is needed [334,336]. It is noted that “efforts to settle on a single, generic version of what constitutes risk communication will be less productive than an open-minded exploration of the multiple forms that comprise today’s vibrant interdisciplinary field” [335] (p. 2240), “Bayesian networks in promoting integrated, inter-disciplinary evaluation of uncertainty in IRBM, as well as the apparent advantages for risk communication with stakeholders, are offset in our case by the cost of obtaining reliable probabilistic data and meta-model validation procedures” [330] (p. 91). It is also noted that adaptive resilience can be achieved even without risk communication if there is a “collective understanding of the system situation” [308] (p. 113).

4. Discussion

A recent scoping review looking at the general engagement with the ‘social’ in the EDMPP academic literature found various problems such as underrepresentation of EDI groups [16]. Our data showed four main findings (a) few to no hits with the term “social risk” were found in the data obtained from most of the search strategies in particular search strategies involving technologies (Table 2 and Table 3), (b) EDI related groups were rarely to not at all mentioned in conjunction with risk in the data investigated (Table 4 and Section 3.2.1–Section 3.2.2), (c) science and technology governance and ethics fields were not mentioned in relation to EDMPPP and risk (Section 3.1.3) and (d) the qualitative analysis (Section 3.2) revealed gaps, for example, technology was only mentioned in a techno-optimistic sentiment and not as a source of risk for EDI groups and EDI covered groups were rarely to not at all engaged with in conjunction with “risk governance”, “risk communication”, ”risk education”, “risk narrative” and “risk perception”. Our data is problematic if looked at through the lens of the risk narrative of the COVID-19 research road map [4], the risk narrative of the UN Office for disaster risk reduction (UNDRR) strategic framework 2022–2025 [15], the discussions around science and technology governance, and the risk narrative through the lens of EDI. We discuss our findings through these four areas below.

4.1. Equity/Equality, Diversity, and Inclusion and Risk Narratives

Equity, diversity, and inclusion (EDI) and other EDI phrases and frameworks [46] are employed as policy concepts to improve research, education, and participation in the workplace for marginalized groups such as women, Indigenous peoples, visible/racialized minorities, disabled people, and LGBTQ2S [46] including those related to environment-focused disciplines and programs [98]. Terms such as equity, equality, inclusion, and diversity are also mentioned in relation to risk narratives [99,100,101,102,103,104,105,106,107,108,109,110,111,112] as are marginalized populations [47,48,49,50]. Many equity, diversity, and inclusion issues were evident in the COVID-19 responses to date [113].

The “UN Research Roadmap for the COVID-19 Recovery” notes over 30 population groups as experiencing highest degree of marginalization [4] (p. 48). EDI frameworks are seen to be useful as tools to deal with various risks such as justice, equity, diversity, and inclusion and to deal with climate change-linked risks [115,124,127] and risk related organizations such as the Risk Management Association have EDI statements [136]. Risk is covered for groups covered under EDI [116,117,118,119,120,122,123,125,126]. At the same time, it is also noted that decreasing EDI efforts carries risks [128,130,131,132,133,134,135].

However, our quantitative and qualitative data suggests a lack of engagement with EDI including the groups covered in the EDI discourses at the nexus of risk and EDMPP. In general, the EDI frameworks seen as useful tools to deal with various risks such as justice, equity, diversity, and inclusion linked to climate change risks [115] and risks for groups covered under [116,117,118,119,120,121,122,123,124,125,126,127] are not engaged with through the lens of EDMPP.

4.2. Risk Narratives and the COVID Research Road Map

The COVID-19 research roadmap [4] mentions risk many times. It highlights the risk of repeating past mistakes by ignoring existing research data or not performing needed research on socio-economic recovery [4]. It is noted that socio-economic recoveries need an understanding of the interconnectiveness of “health, the environment, politics, economics and societal well-being” [4] (p. 30) and the report highlighted the need to provide data that enables shared risk and does not burden marginalized groups the most [4]. Our data suggest that the research reality around the literature we looked at fits the category of “not performing needed research” [4]. Many research priorities are linked to the concept of risks often in conjunction with marginalized groups {RP3.1.2], [RP1.1.3], [RP1.4.1], [RP1.4.2], [RP1.5.3], [RP2.4], [RP2.4.1], [RP2.4.4], [RP2.4.3], [RP2.4.2], [RP3.5.1] and [RP2.5.2]. Our study provides empirical data that research on new global governance structures that takes into account local realities, exhibits long term thinking and ensures the so far rare reality of the meaningful participation of marginalized groups [4] is missing. If one would use the list of marginalized groups identified in the UN Research Roadmap for the COVID-19 Recovery document [4] (p. 48) the neglect of the groups in the list we did not cover would very likely also be found in the literature we covered. Similarily, the data identifies that “to build fit-for-purpose global governance structures that enable effective responses to transnational risks while also encouraging everyone to act in solidarity in the face of shared responsibilities [RP1.4.3]” [4] (p. 30) is still missing.

4.3. Risk Narrative and the UN Office for Disaster Risk Reduction (UNDRR) Strategic Framework 2022–2025

The UNDRR as the very name indicates, focuses on risk [15]. In the UNDRR strategic framework 2022–2025 it is flagged that risk informed “evidence, innovation, and good practices on risk” that also includes systemic risk has to be used to inform decision-making processes is needed [15] (p. 4). The report flags the importance of evidence-based advocacy to mobilize citizens, risk knowledge and information, monitoring of action results [15]. Our data suggests that these goals are not met, and that the “science-policy practice interface” [15] (p. 10) highlighted as essential is very weak in relation to risk and marginalized groups as is the goal of “integrating human rights, gender equality and the rights of persons with disabilities into disaster risk reduction” [15] (p. 16). Our data suggests that all strategic objectives are so far not supported by the risk and EDMPP research as it pertains to marginalized groups (strategic objectives 1: the generation of quality risk information and analysis which includes systemic and cascading risk (deliverable 1.1.3), scientific research agenda on disaster risk reduction (deliverable 1.1.4), applying to action of disaster risk reduction practices, knowledge and methodologies (deliverable 1.2.1), monitoring based on data disaggregation by sex age and disability (deliverable 1.3.1), to develop and implement multi-stakeholder, gender responsive, inclusive risk governance (result 2.2 and deliverable 2.2.1) and strategic objective 4: “Mobilize governments and other stakeholders through advocacy and knowledge sharing to make disaster risk reduction central to sustainable development” [15] (p. 24). Our study suggests that data is still not generated in the academic literature for the deliverables mentioned and problems flagged.

4.4. Science and Technology and Risk Narratives

Perception of risk is a social and cultural construct and is impacted by many factors [45]. Risk perception and an understanding of risk narratives of scientific and technological advancements is a critical aspect of scientific literacy [51,52,53,54,55,56]. However, risk perceptions are shaped by the values of individuals and groups [57,58,59,60] and are influenced by motivated reasoning [57,58,59,60]. Risk narratives are mentioned extensively in conjunction with science and technology governance and various ethics discourses including environmental ethics [82,83,84,85,86,87,88,89,90,91,92,93,94,95,96]. Various studies describe the problems of risk narratives in relation to technologies such as artificial intelligence, machine learning, machine reasoning, decision support systems, virtual coaches, conversational agents, Bayesian network approaches [61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81]. Risk narratives are mentioned extensively in conjunction with science and technology governance and various ethics discourses including environmental ethics [82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97]. None of the science and technology governance terms and ethics field are present (Section 3.1.3). Furthermore, EDI groups are rarely mentioned in conjunction with risk narratives (Table 4), and within the abstracts used for qualitative analysis technologies were only mentioned with a positive angle in relation to EDI if mentioned (Section 3.2). Our findings pose problem for performing high-quality risk governance in relation to technologies used in EDMPP. This is of specific importance in relation to marginalized groups if the purpose of risk governance is to provide “a conceptual as well as normative basis for how to deal responsibly with uncertain, complex, and/or ambiguous risks in particular” [40] (p. 431) and to inform policy [337]. Our data also suggests that risk literacy and risk education around marginalized groups has not improved if indeed they are a topic of investigation to start with. However, high risk literacy and risk education leading to risk literacy [42] is seen as needed to improve risk governance [41]. Our data suggest that all three components of risk literacy “(1) understanding of the state of risk, (2) understanding of risk perception, and (3) understanding and practicing the risk coping (including both of risk management and risk communication)” [41] (p. 1260) are missing in the academic literature around EDMPP and risk in relation to technologies and marginalized groups. This area is also neglected in the science and technology governance literature. The National Academy of Medicine’s Committee on Emerging Science, Technology, and Innovation in health and medicine (CESTI) proposed a technology impact and governance framework [97] which includes principles of relevance to EDI and risk narratives in relation to EDMPP and the generation of technologies and algorithms for use in EDMPP. These examples of the principles include fairness, autonomy/individual and group self-determination, collective good with concept such as solidarity, civic responsibility and stewardship and individual good whereby risk benefit is linked to the principle of justice, collective and individual goods [97] (p. figure page 45). The lack of risk narrative data in the literature we covered in relation to marginalized groups is problematic to fulfill the principles of the CESTI framework. Our data also shows problems for the expected cross sectors and cross discipline efforts and the efforts to involve stakeholders reflecting many different groups expected for technology governance [97] given that technology is an important part of EDMPP. It is argued that “The extent to which a technology’s benefits are maximized and risks mitigated (and how benefits and risks are defined) often depends less on explicit ethical principles and values guiding the work itself, and more on the policies, norms, standards, and incentives of the particular sector that shapes a technology’s development and deployment” [97] (p. 41). Our data suggests that there might be problems with how benefits and risks are defined and that there is a lack of information generated for “the policies, norms, standards, and incentives of the particular sector that shapes a technology’s development and deployment” [97] (p. 41). There is a lack of engagement with risk and marginalized groups and EDMPP, especially in conjunction with technologies. Our study suggests also that data is not generated that could be used to generate the plausible future scenarios seen as needed [97] and the unearthing of potential conflicts [97] including social conflicts between individuals and groups.

4.5. Limitations

The search was limited to abstracts in selected databases and English language literature. As such, the findings are not to be generalized, which was also not the purpose of the study, to the whole academic literature, non-academic literature, or non-English literature. The hit counts produced are based on the co-occurrence of terms and do not indicate whether the content is relevant to risk narratives of EDMPP and are a maximum and do not account for duplicates between databases and within abstract. We also chose to search the data obtained with the search strategies for certain terms for example depicting ‘the social’ and EDI groups but our terms used are not exhaustive. Although this study has various limitations, the findings allow for conclusions to be made within the parameters of the searches and the character of the analysis.

5. Conclusions and Future Direction

Our data showed four main findings (a) few to no hits with the term “social risk” were found in the data obtained from most of the search strategies in particular search strategies involving technologies (Table 2 and Table 3), (b) EDI related groups were rarely to not at all mentioned in conjunction with risk in the data investigated (Table 4 and Section 3.2.1–Section 3.2.2), (c) science and technology governance and ethics fields were not mentioned in relation to EDMPPP and risk (Section 3.1.3) and (d) the qualitative analysis (Section 3.2) revealed gaps, for example, technology was only mentioned in a techno-optimistic sentiment and not as a source of risk for EDI groups and EDI covered groups were rarely to not at all engaged with in conjunction with “risk governance”, “risk communication”, ”risk education”, “risk narrative” and “risk perception”.

Our data is problematic if looked at through the lens of the risk narratives of the COVID-19 research road map [4], the risk narrative of the UN Office for disaster risk reduction (UNDRR) strategic framework 2022–2025 [15] and the discussions around science and technology governance. It is also suggesting that existing EDI strategies so far did not lead to research questions of relevance to groups covered under EDI and to tackle the problems EDI groups face within EDMPP. As to future research, numerous studies can be performed with various stakeholders to better understand and fix the gaps we found. To name two avenues of research: given that EDI is increasingly employed as a prerequisite for grant proposals one can investigate whether the wordings really trigger research related to EDI groups or only lead to what we call “bureaucratical EDI” the accessibility to do research. Another research avenue could be to develop self assessment tools for people involved in EDMPP and risk especially the ones employing technologies such as machine reasoning to increase their awareness of their level of knowledge around EDMPP and social risks especially of marginalized groups so they can consciously avoid designing or deploying biased algorithms and other products.